Re: [RAM] Re: the separation of ID/RLOC

[Tony Li <tli at cisco.com>]

On Jul 10, 2007, at 1:25 AM, Christian Vogt wrote:

> > > if you want to an enable edge network to decide through which  
> > > provider
> > > ingress packets are received, then the edge network must be able  
> > > to tell
> > > remote edge networks one specific transit locator to which they  
> > > should
> > > send packets.
> >
> > A destination can certainly express a preference, but as always, the
> > holder of the packet gets to decide on the next hop.  This may or may
> > not include dropping the packet.  This is best effort and there  
> > are no
> > guarantees.
>
> before I reply to your email, let me make sure there is no  
> misunderstanding.
>
> What I was thinking is that, if edge network E communicates with edge
> network F, and E directly connects to providers P1 and P2, then E  
> should
> be able to choose one of P1 and P2 for (1) egress packets that E sends
> to F, and (2) ingress packets that E receives from F.
>
> I think that we actually agree on this, right?


No, sorry.  E gets to decide which egress it uses.  However, F also  
gets to decide what egress it uses.  This is necessary for symmetry.   
Since F and E may not agree, ingress policy may not be respected.

If E prefers P1 as ingress and F prefers P2 as egress and there's no  
P1-P2 link, then you end up violating E's preferences.

The only way that E can guarantee its ingress policy is to not  
advertise reachability through sub-optimal providers.

Tony

_______________________________________________
RAM mailing list
RAM at iab.org
https://www1.ietf.org/mailman/listinfo/ram